Modelling study

The model to accompany the study will evolve from the mathematical description of the important points in the model, such as the location of the joints, the hand, the target, and the dependencies between them. The arm will consist of three sections: The arm, forearm and hand.

Once the correctness of the underlying code has been established, the model will be extended with a visual representation of the arm, consisting of a wire frame drawing of each component. Following this, a simple language will be implemented to allow easy control of the arm, both for direct manipulation, and animated behaviour.

Once this is complete various control methods can be implemented to operate the arm automatically. For this to work there will need to be a way of selecting the method to use, some way to establish how close the arm is to the target and some way of evaluating its performance. The distance of the arm from the target is given by Pythagoras: D = sqrt((x1-x2)² + (y1-y2)²). The performance can be evaluated in a number of ways, but the most effective (and simple) seems to be to count the number of individual movements required from the start position to reach the target.

Following of from that, the different algorithms will be tested under more stringent conditions, such as having blocking objects obstructing the normal path of the arm, and limiting the degrees and extents of freedom available to the arm. One such limitation is the control of a human arm, giving limits to the rotation of all the joints in the arm.

Other extensions could include a ‘graph of progress’ allowing a more refined evaluation of the algorithms for arm movement, enhanced visual drawing of the arm, click and drag repositioning of the joints, although owing to time constraints most of these will be unrealisable.